The dopaminergic system plays an important role in the control of a variety of functions including motor activity, cognition, emotion, positive reinforcement, pleasure and reward, food intake, and endocrine regulation. It is implicated in such diseases as schizophrenia (excess of dopamine), Parkinson's disease (lack of dopamine), drug abuse, and alcoholism, all of which involve imbalances in the level of dopamine. Five subtypes of dopamine receptors mediate the function and the level of the dopamine neurotransmitter. Dopamine receptor antagonists have been developed to block hallucinations and delusions that occur in schizophrenic patients, whereas dopamine receptor agonists are effective in alleviating the hypokinesia of Parkinson's disease. However, blocking dopamine receptors can induce side effects similar to those resulting from dopamine depletion, and high doses of dopamine agonists can cause psychoses. Thus it is important to develop agonists and antagonists that are selective for a particular receptor subtype. Because of the similarity in the binding sites for these receptors, developing such subtype specific receptors has been too slow, partly because of the lack of 3D structures for any dopamine receptor or any homologous receptor Dopamine receptors belong to the large G-Protein Coupled Receptor (GPCR) family of seven helical transmembrane proteins. It has not yet been possible to obtain experimental 3D structures for these receptors, but we have developed the MembStruk computational strategy to predict the 3D structure sufficiently accurately that the HierDock computational strategy can predict the binding sites for small molecules. Indeed preliminary studies for bovine rhodopsin, human D2 dopamine receptor, p2-adrenergic receptor, and olfactory receptors are all consistent with available experimental results. We find that the binding site of dopamine in human D2 receptor, and epinephrine in (32 adrenergic receptor is in excellent agreement with the mutation experimental results. Here we propose a rigorous validation of these computational methods by predicting the structure and ligand binding sites of various agonists to both D1 and D5 subtypes of dopamine receptor. The predictions made by computations will be validated and tested by experiments. This exploratory collaborative study between experiments and theory will be used to validate the ab initio computational methods and will lay the basis for testing the feasibility of using these first principles methods for structure-based design of anti-Parkinsonian and anti-schizophrenic drugs, specific to each dopamine receptor. Moreover these computational methods developed and refined in this project can be applied for other more complex GPCR systems like the lipid and peptide receptors. [unreadable] [unreadable]